Sunday, August 09, 2009

Position sizing

Most of us enter the stock market burdened with faulty biases and misguided conceptions about how to succeed in our trading program. We think that we can beat the market by being smarter than the market. We think that we will establish a winning method of stock selection by way of superior information, that the road to profits will be paved by our consistent ability to pick winners. In short, we enter the market with an emotional commitment to our own judgement. Unfortunately, it is this highly individualized bias that is our ultimate nemesis. Experienced traders know that the key to market success has less to do with a uniquely developed style and method of picking stocks, and more to do with a disciplined execution of sensible and rigid money management principles.


These “judgement biases” are described meaningfully by Van K. Tharp, in his significant contribution to the business of trading, Trade your Way to Financial Freedom (McGraw-Hill,1999). Our trading systems are handicapped by the fact that we typically trade our beliefs about the market, we accept conventional representations of information, and adopt trading practices that undermine trading success by inadequately protecting against actual risk. Further, we consistently and foolishly bet against the odds and succumb to undisciplined approaches to trading.


What precisely are we talking about here? Let us assume that a trader has developed a trading system that generates a positive expectancy – a profitable trading strategy that either has a superior winning percentage (% of winning trades) or a superior profit factor (where the profits of winning trades far exceeds the losses of losing trades). There remains a variable that will, in fact, determine the actual long-term profitability and risk factor that results: the position size.


An often-cited experiment by Ralph Vince, who has been an seminal author of important trading analysis treatises (Portfolio Management Formulas and The Mathematics of Money Management), tested the gambling performance of 40 Ph.D’s on a simple computer game with a 60% chance of winning. Each were given $1,000 in play money and instructed to bet however much money they wished (or could) over 100 trials. The end result was that only two made money.


The undoing of this group of individuals, in fact, is the undoing of many traders: we tend to bet more when we think we are going to be right (or need to be right), and bet less when we are less certain of the outcome. In other words, our emotions have dictated our risk. If the 40 Ph.D’s had religiously accepted the mathematical probability that a constant bet of $10 over 100 turns would have resulted in a 20% gain. Yet, these individuals were inclined to vary their bets (risk) in hopes of achieving better results.


There are varying methods of determining what is an optimal position size. Some are related to drawdown. Some are related to other statistical variables that a given trading system produces. But for general purposes we can simply recognize some basic common sense principles. First, in order to trade properly one must remove emotional attachment to a position. The only way this can be done is if there is enough available trading capital to allow for actual risk. You cannot trade effectively with inadequate capital. Second, traders must limit risk in a quantifiable fashion. For the purposes of retail traders like the majority of Stock Trends followers, a benchmark of 1-2% of available trading equity should be at risk in each trading position (Where the risk is defined by a quantifiable measure. For example, the average loss per trade plus one standard deviation.)
Obviously, position size will be determined by available capital, so the money management guidelines will be predicated by sufficient funds available. Finally, constant position sizing will enable a trading system to minimize the hazardous effects of our inherent, emotionally charged biases.

The best analysis tools I’ve seen that provides position size guidance is the m3 Money Management Modeler, designed by trader Brian Ault of Fulcrum Shift Trading. Brian’s platform looks at trading as a probability assignment, taking essential input variables to determine HOW MUCH to trade on given positions – either stocks or options. Traders, advanced and novice, should learn how to incorporate this kind of risk analysis in their trading strategies. A good trading plan should factor  in probability analysis.

[Most of this is post is a reprint of a Stock Trends Weekly Reporter editorial from 2003]

Post a Comment